Skip to content

Latest commit

 

History

History
46 lines (29 loc) · 2.38 KB

README.md

File metadata and controls

46 lines (29 loc) · 2.38 KB
Raw

Stimulus domain transfer in recurrent models for large scale cortical population prediction on video (Code)

Code to reproduce results of the NIPS 2018 paper: "Stimulus domain transfer in recurrent models for large scale cortical population prediction on video".

Requirements

  • docker, nvidia-docker (version 1), and nvidia-docker-compose. You can easily run it with nvidia-docker version 2. In that case have a look at the nvidia-docker-compose.yml.jinja and extract the options for the notebook service to start the container.
  • GIN along with git and git-annex to download the data.

Data License

The data shared with this code is licensed under a This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. This license requires that you contact us before you use the data in your own research. In particular, this means that you have to ask for permission if you intend to publish a new analysis performed with this data (no derivative works-clause).

Creative Commons License

Quickstart

Go to a folder of you choice and type the following commands in a shell of your choice:

git clone https://github.com/sinzlab/Sinz2018_NIPS.git
cd Sinz2018_NIPS

# get the data
gin get cajal/Sinz2018_NIPS_data # might take a while; fast internet recommended

# create a file with DB credentials
echo "DJ_HOST=134.2.118.234" >> .env
echo "DJ_USER=public_access" >> .env
echo "DJ_PASS=ffeWQ@7O" >> .env

# create docker container (possibly you need sudo)
nvidia-docker-compose -t docker-compose.yml.jinja build notebook0

Then you can start the container via

nvidia-docker-compose -t docker-compose.yml.jinja up notebook0

Now you should be able to access the jupyter notebooks via YOURCOMPUTER:2018 in the browser.

Custom Database Server

You can also run the notebook with your own database server. In that case you need to insert the content of Sinz2018_NIPS_data/dbdump/nips2018.sql into your own database and change the DJ_HOST, DJ_USER, DJ_PASS parameters accordingly.